The atlantoaxial synovial folds have been described in detail as physiological anatomical structures in human medicine, providing insights into their morphology, function, and potential involvement in clinically relevant disorders (1–3). In humans, the atlantoaxial joint is composed of three distinct joint cavities, the median, left and right atlantoaxial joint cavities, which sometimes intercommunicate. Additionally, the Bursa atlantodentalis, located between the Dens axis and the Ligamentum transversum atlantis sometimes communicates with the median atlantoaxial joint cavity (4).

The atlantoaxial and atlantooccipital joints in dogs form one large synovial cavity, which is subdivided into a pair of lateral atlantooccipital cavities, a median cavity that runs between the dens and the atlas, and in addition the paired left and right atlantoaxial joint cavities (5). This means, that in dogs, the atlantoaxial joint cavities intercommunicate and communicate with the atlantooccipital joint cavities, and this intercommunication sometimes also occurs in humans (5).

In general the joint capsule consists of an outer fibrous layer (Stratum fibrosum), and an inner synovial layer (Stratum synoviale), the latter can be further subdivided in intima and subintima (Figure 1) (4). The synovial layer forms physiological protrusions or synovial folds at various points of the joint capsule, some of which protrude into the intraarticular space (Figure 1). The intima consists of epithelial-like cells, the synoviocytes and the subintima, also referred to as subintimal stroma which fills the space between the synoviocytes and the outer fibrous layer. The synoviocytes form a discontinuous cell layer, and lack a basement membrane (4). Two types of synoviocytes can be distinguished; macrophage-like cells and fibroblast-like cells, the existence of both cell types was confirmed in humans and dogs amongst other species (6). The synoviocytes are arranged in layers of one to three cells. A differentiation between both types of synoviocytes is limited using light microscopy and requires immunohistochemistry and/or transmission electron microscopy (6). The subintima merges with the outer fibrous layer of the joint capsule. It contains regularly blood vessels, nerve fibers, lymph vessels and might contain some elastic fibers (1).

According to tissue composition and shape, three distinct types of subintimal stroma are defined: areolar, fibrous and adipose types. The areolar type is composed of loose connective tissue and shows extensive vasculature. It is further characterized by the formation of villous protrusions. The fibrous type shows dense, fibrous connective tissue and poor vascular supply. The subintima of the adipose type consists mainly of fat cells (6). There are also mixed forms of various characteristic types, such as the fibroadipose type.

In humans, the synovial folds are located in the ventral and dorsal aspect of the right and left lateral atlantoaxial joints, respectively (7, 8). These crescent-shaped intra-articular projections originate from the synovial layer and their first description is attributed to Henle in 1855 (3, 9). The folds are subdivided into 3 parts based on their appearance: the base lies directly on the fibrous layer of the capsule, seamlessly joins the middle part, and tapers to an apex that extends into the atlantoaxial joint (Figure 1) (10). Synovial fold volume varies with size and gender, being larger in males (8, 11, 12). While the children population shows a higher number of larger folds than adults, there is no correlation between age and synovial fold volume in adults (2, 11). Ventral synovial folds have been described in human literature to be significantly larger than dorsal folds (11, 12). Adipose synovial folds are commonly found in children and also in the atlantooccipital joint of adults, whereas fibrous or fibroadipose synovial folds are typically observed in the atlantoaxial joint (2, 3). The nerve fibers within the subintimal stroma of the folds are found both around blood vessels and as free fibers, highlighting their roles in vasoregulation and nociception (13), with the ventral rami of the C1 and C2 spinal nerve innervating the atlantoaxial joint capsule (14).

The function of synovial folds is still unclear. They are thought to be “passive space fillers” that fill the noncongruent aspect of the joint and adapt to joint motion and are therefore often called “meniscoids” (3, 7). They therefore act as a protective shield and lubricant for the articular surfaces (3, 7, 9, 15). In humans, several hypotheses have been proposed regarding the potential involvement of synovial folds in disease processes and their clinical relevance. These include synovial fold entrapment and extrapment. Synovial fold entrapment involves the fold being caught between the articular surfaces (10, 16), while extrapment refers to the fold being trapped outside the articulation, between the joint capsule and the vertebra, due to abnormal movement (17, 18). Both conditions are believed to contribute to the development of torticollis (2, 19–21). Additionally, joint immobility has been hypothesized to lead to fibrosis of the synovial folds, which could reduce the range of motion (18). However, the majority of studies on atlantoaxial synovial folds in humans have focused on their role in whiplash injuries (1, 22–26). Autopsy studies have reported contusions or tears of the synovial folds in whiplash patients, which may contribute to cervical pain (22–26) and balance and visual disturbances have been reported to accompany upper cervical pain and are of particular clinical importance in the context of whiplash (27–29).

Given the potential association of synovial fold injury with neck pain in humans, there is growing interest in in vivo imaging techniques. Synovial folds were first successfully visualized in vivo on MRI using an isotropic three-dimensional T1-weighted spoiled gradient echo volumetric interpolated breath-hold examination fat-suppressed (VIBE) sequence and a T2-weighted sampling perfection with amplification-optimized contrast using different angle evolutions (SPACE) sequence (30). Recent work has attempted to correlate the histological results and MRI findings of synovial folds. Based on signal intensities on VIBE and SPACE sequences, synovial folds could be classified as containing mostly adipose tissue (hyperintense on VIBE and SPACE), mostly connective tissue (hypointense on VIBE and SPACE), or mixed tissue (mixed hyper- and hypointense on VIBE and SPACE sequences), showing a good correlation with histological findings (31). Another method using in-phase, out-of-phase, fat-only, and water-only Dixon MRI images improves agreement with histological findings to a “substantial” level (18, 32).

Despite the wealth of information in human anatomy, the existence, morphology and potential function of synovial folds within the normal canine atlantoaxial joint remains unexplored, with only one mention of synovial protrusions within the atlantoaxial joint in the veterinary anatomy literature (33). Our aim was to provide the first comparative description of synovial folds in the presumably normal canine atlantoaxial joint using advanced MRI imaging techniques, gross anatomical examination, and histological analysis.

Our findings demonstrate that synovial fold are consistent anatomical components of the canine atlantoaxial joint, with morphology and tissue composition comparable to those described in humans, and highlights the potential for high-field MRI—particularly fat-suppressed VIBE sequences—to visualize these structures.

Synovial folds of the canine atlantoaxial joint were visible on MRI in dog cadavers without known atlantoaxial disease, with the best visualization achieved using the VIBE sequence. In this sequence, fat and fibrous tissue- the primary components of synovial folds- appear hypointense, whereas the surrounding structures such as synovial fluid are isointense to the musculature, leading to an increased contrast between the synovial folds and the surrounding structures. Furthermore, the VIBE sequence provides high spatial resolution three dimensional images, which facilitates the visibility of small anatomical structures. Additionally, imaging was performed using a 3 Tesla MRI system, which offers high spatial resolution and likely contributed to the clear visualization of these small structures. Although synovial folds are known to be visible on 1.5 Tesla MRI in humans, it remains uncertain whether similar visibility could be achieved in dogs—particularly given their generally smaller size—or when using lower-field MRI systems.

In all dogs examined, ventral synovial folds were identified by MRI and confirmed by anatomical dissection. While dorsal synovial folds were visible on MRI only in three dogs, their presence was confirmed in all dissected dogs.

In all dogs examined, dorsal synovial folds were smaller than the ventral synovial folds, which is also consistent with previous findings in humans (11, 12).

The lack of visible dorsal synovial folds on MRI in the remaining 3 dogs was likely attributable to their small size and individual anatomical variation, particularly related to overall body size. This became obvious as dorsal synovial folds were clearly delineated in larger dogs (dog 1, Shar Pei and dog 5, Giant Schnauzer). Notably, dog 5, the largest, had also the highest body weight. However, dog 3, a mixed-breed dog, despite weighing more than dog 1, did not exhibit clearly delineable dorsal synovial folds (Figure 9). This discrepancy might be breed-specific, related to different composition of the folds or attributed to age differences as described in humans, were older individuals tend to show smaller synovial folds (7), similarly dog 1 was 4.4 years old, while dog 3 was 6.5 years old. The youngest dog in this study, a 4-month-old puppy, had large and well-defined dorsal and ventral synovial folds (Figure 9). This observation suggests that, similar to observations in humans, synovial folds may also be larger in young than in adult dogs (2, 11). In some dogs, ill-defined fat-suppressed, VIBE strongly hypointense structures were visible at the expected location of the dorsal synovial folds, suggesting their presence, which was confirmed in anatomical dissection in all but one dog, where it was confirmed on histology.

In humans, the synovial folds are also believed to be more adipose in composition at a young age (2).

The above-described individual factors, combined with the limited spatial resolution of MRI, may have contributed to the inability to consistently visualize these small structures.

One limitation of the present study is that all examinations were performed on cadaveric specimens. Although a high-resolution protocol using a 3 Tesla MRI system was employed, it remains possible that post-mortem changes, including those related to prior freezing and thawing, may have introduced artifacts—such as abnormal positioning or altered signal intensity of the synovial folds. Additionally, the absence of normal perfusion may lead to a signal alteration. Contrast-enhanced imaging with gadolinium, which is routinely used in clinical veterinary MRI, was not performed. As a result, we were unable to evaluate the behavior or enhancement pattern of synovial folds following contrast administration.

To better characterize these structures, this study provides a detailed anatomical description of the shape and architecture of the larger ventral synovial folds, which were consistently visible on MRI.

Complementary histological analysis further elucidated the morphology and tissue composition of the synovial folds at the cellular level.

Notably, the shape of the canine synovial fold, especially the ventral fold permitted subdivision into three parts: (1) base, (2) middle, (3) apex, corresponding to the observation in humans (10). In addition, we were able to assign the three types of subintimal stroma described in humans—areolar, fibrous and adipose—to the canine synovial folds on the basis of the tissue composition (6). The synovial folds in this study were described at different levels of the atlantoaxial joint and, interestingly, the composition of the subintimal stroma varied depending on the location in the joint. At the level of the dens axis the subintima of the ventral synovial fold showed characteristics of the fibrous type with densely packed fibrous strands and poor vasculature. Although the subintima of the large ventral fold in the area lateral to the dens axis had fibrous strands at the base which were more loosely packed and more vascularized than the fibers of the fold at the level of the dens axis, the middle part of the fold in particular consisted of densely packed, almost axially arranged fibers representing the fibrous type. Sections of the ventral fold at a more lateral level and in the periphery of the atlantoaxial joint exhibited features consistent with the adipose type. Studies of synovial folds in humans have also described mixed forms of the various characteristic types, such as the fibroadipose type, and the subintimal stroma also regularly contains blood vessels, nerve fibers, lymphatic vessels and elastic fibers (1). However, it is not yet clear from the literature whether the mixed types are observed in humans at specific localizations or whether—as in our case—the composition of the tissue changes over the course of the synovial fold.

In addition to the presence of dorsal synovial elevations, which could be classified as areolar due to the loose, well-vascularized stroma, a comparatively longer dorsal synovial protrusion was observed. This protrusion had a base consisting of densely packed fibrous strands and was intermixed with fat cells and blood vessels in the middle portion. Consequently, this fold corresponded to a fibroadipose subintimal type.

Altogether, the histological observations of the atlantoaxial joint’s synovial folds in that dog correspond to the findings in humans, where fibrous and fibroadipose synovial folds are typically observed in the atlantoaxial joint (6). The occurrence of different tissue composition and shapes allows for the hypothesis that different mechanical stresses act on the folds. Smaller elevations, protrusions and folds presumably serve to increase the surface area of the synovial layer, which also represents the site of production for synovial fluid.

In human medicine, synovial folds are hypothesized to function analogously to menisci, contributing to joint congruity and stability, protecting and lubricating the articular surface, assisting in weight-bearing, and dissipating mechanical stress (3). Our investigations have demonstrated that the largest synovial folds protrude into the joint space, i.e., between the articulating surfaces. This suggests an important function in maintaining joint congruity and/or enhancing the weight dissipation between the articulating bones.

In addition, we also observed many smaller protrusions of the synovial layer, without reaching into the joint space. Here, their function is probably more to increase the surface area and therefore support lubrication.

Nevertheless, it is not possible to draw definitive conclusions about the general histological structure of the synovial folds from one dog. Further research based on more animals is required to elucidate the differences in size and composition of synovial folds between age groups and breeds in dogs.

In humans, the clinical relevance of synovial folds stems from the presence of nociceptive nerve fibers. Additionally, there is ongoing debate regarding whether synovial folds contribute to proprioceptive feedback for sensorimotor control, potentially establishing neurophysiological connections to the vestibular and visual systems via cervical proprioceptors (34, 35). However, with the methods we used on the one specimen, we could not clearly identify nerve fibers. Further investigations including specific histological staining or immunohistochemistry are required.

Beyond histology, the tissue composition of synovial folds, as evaluated via MRI, has attracted growing clinical interest—particularly in human medicine. Current research is investigating correlations between MRI signal characteristics of synovial folds on specific sequences (e.g., Dixon, VIBE, and SPACE) and histological findings (18, 31, 32). In our study, the ventral folds of dog 1 exhibited homogeneous strongly hypointense signal on the VIBE sequence by MRI imaging, while the central portion appeared partially hyperintense on both T1w and SPACE sequences. This imaging pattern suggested the presence of central adipose tissue. Further, the peripheral regions of the ventral folds were hypointense on T1w and SPACE sequences in dog 1, which is indicative of fibrous tissue. These MRI imaging findings were consistent with the histological results from dog 6, where the synovial folds were found to have a fibroadipose composition. This concurrence suggests that it may be possible to determine the composition of the synovial folds in dogs based on its MRI signal in different sequences as described in humans (18, 31, 32).

The potential clinical utility of such imaging assessments is highlighted by a recent MRI study comparing synovial folds in whiplash-injured people with those in an age-matched control group, showing that the lateral atlantoaxial synovial folds in the whiplash group had an increased fibrous component and were shorter, which suggests a composition alteration and could be of clinical significance in cases of chronic whiplash-associated disorders (1, 8).

Although whiplash injuries are less common in dogs, the anatomical structure of the synovial folds indicates a similarity to humans, suggesting a similar function, and that certain diseases or mechanical impacts can affect the synovial folds of the dog in a similar way or that similar adaptation mechanisms can occur. For example, small breed dogs often suffer from atlantoaxial instability (36), which could lead to increased pressure on the synovial folds due to increased movement or abnormal positioning. Furthermore, the use of collars in dogs may exert additional pressure on these structures, potentially leading to chronic changes.

Our study offers novel insights into veterinary medicine by being the first to document the presence of synovial folds in the atlantoaxial joint of dogs on 3 T MRI, being best visible on VIBE sequence. We provide detailed information on their localization, size and tissue composition. Thus, this study contributes to a better understanding of canine synovial folds. Future research is needed to investigate the morphology and MRI characteristics of atlantoaxial synovial folds in dogs of different ages, different clinical histories and different breeds to shed light on their anatomical variability and possible involvement in clinically significant disorders.